Folks; After some analysis of the PMDG 737 fuel system, I make the following observations and/or recommendations: The fuel selector is being modulated back/forth between right and left tank internally by the panel. As a result, fuel burn does not reflect any differential in fuel flow if one engine is burning more than the other, i.e. with one pulled back at idle for a malfunction or simulated single engine approach. Aditionally, if one tank reaches zero quantity--even with the crossfeed open--both engines flame out because internally they are both being fed by the one (empty) side. With crossfeed closed, and a single center boost pump on, the center tank erroneously supplies fuel under pressure to *both* engines. This is related to using the single fuel selector rather than setting the TURB_ENGINE_x_FUEL_SELECTOR values independently. Additionally, because the selector is set to center tank, both engines flame out when center tank quantity reaches zero. With all sources of boost pump pressure turned off on one side and crossfeed closed, the appropriate engine fuel press light does not illuminate. In my own panel writing, I have taken to working fuel system logic independently of MSFS. I take regular snapshots of engine/APU fuel flow (every 200 msec works well) and accumulate fractional amounts attributed to each fuel tank providing fuel to that engine during that snapshot. Every 5 sec I take the total amount of fuel consumed (using MSFS total fuel quantities) since the last 5 sec update, and decrement the quantity (kept in a panel local variable) of each tank in proportion to that tank's relative contribution to total fuel used during the last 5 sec. Once per minute, I write back the locally maintained quantities to each tank in MSFS to keep the weight and balance correct, and if the MSFS quantity increases (or drops by more than ~1000 lbs), I reset the local variables assuming the fuel quantity was modified by the user. The panel itself only sees the local variables for quantity. I have implemented a working fuel system panel for a complex 7-tank B747-200 fuel system in this manner...a 3-tank 737 should be fairly straightforward. Anyway, this is just one way that a good panel programmer can operate beyond the constraints of the MSFS fuel logic.RegardsBob ScottATP IMEL Gulfstream II-III-IV-V

I'm with you on this one Bob.I find a longer cross-country difficult to do. Also, the APU will startup in flight on its own.Jeff

Jeff, I agree with you about the APU, i can't seem to shut it off either! CheersJamie Blethyn, UK